A proposta básica desse estudo é definir malhas ótimas de sondagem a serem aplicadas em pesquisas futuras de jazidas de minério de ferro. Realizou-se um estudo geoestatístico utilizando-se como parâmetro de otimização, a distribuição das classes dos recursos, obtidas a partir do Índice de Risco (IR), adotado como padrão, na avaliação de depósitos e minas de ferro pela Vale. Esse índice baseia-se no emprego da krigagem da indicatriz de minério para classificação dos recursos, avaliando-se, então, a influência da variação da malha de sondagem na categorização dos mesmos, pela criação de malhas regulares de sondagem "virtuais" com espaçamentos diversos. Estudos comparativos foram realizados em minas existentes e em projetos, tanto no Quadrilátero Ferrífero, quanto em Carajás, obtendo-se malhas de sondagem indicativas por província, sendo que, para novas áreas, criou-se uma metodologia de pesquisa sequencial, com adensamento gradativo da malha de sondagem e avaliação dos riscos em cada etapa.

The weathering of Paleoproterozoic itabirites (metamorphic-banded iron formations) and dolomites from the Cauê and Gandarela Formations in the Quadrilátero Ferrífero (QF), Brazil, produces supergene iron ore with different mineralogical, chemical, and physical properties. In this work, we present a methodology to assess the changes in chemical and physical features of those rocks during weathering, via quantitative analyses of mineral assemblages. These mineral assemblages were calculated from chemical analyses of fresh and weathered samples collected from drill holes drilled in different iron ore deposits in the QF. In general, the number of mineral species found in fresh or/and weathered itabirite is restricted, which helps the quantification of the mass and volumes of minerals by normative calculation in a large dataset of drilling and channel samples. The calculation of the bulk density takes into consideration, besides the mineral phases, the voids and free water in the altered rock matrix. This study shows that the estimated porosity in supergene ore varies from 0% to 20%, for compact materials, and from 15% to 55% for friable rocks, indicating an important process of rock matrix dissolution during the weathering of itabirites. In this process, MgO, CaO, and FeO are leached out from carbonates, talc, and amphiboles. Magnetite is oxidized to hematite, releasing Fe²⁺, which is oxidized and precipitates as Fe-hydroxide. There is a concentration of Fe₂O₃, MnO, Al₂O₃, and SiO₂ in the supergene ore (saprolite) by residual enrichment or recrystallization of hematite, goethite, quartz, manganese oxides, and kaolinite. A calculation of weathering effects on the original protoliths allowed for the establishment of a correlation between different types of fresh itabirites and their corresponding weathered materials. The calculation was carried out in several steps, to account for changes in porosity and masses and has taken into consideration differences in the mineralogical composition of the protolith. Within the weathered zones, a strong link is observed between the existence of collapse on the topographic surface and the presence of supergene ore underneath. The partial to total dissolution of quartz and carbonates from the protolith itabirite results in very porous materials and leads to gravitational collapses.

Combinações de minério provenientes de diferentes frentes de lavra e pilhas de homogeneização são, comumente, empregadas para reduzir a variabilidade dos teores nas plantas de beneficiamento. A metodologia proposta, nesse trabalho, visa a quantificar a variabilidade das diversas etapas em um sistema de homogeneização. O parâmetro químico avaliado é sílica (SiO2) de minério de ferro, na faixa granulométrica inferior a 0,15 mm. Este é um contaminante errático, nesse tipo de depósito. Os resultados mostram a variância esperada para os diversos tamanhos de pilhas. Fica demonstrado que a variância, reduz com o aumento da massa da pilha.

Moisture is a critical variable in iron-ore processing, handling and transportation. During beneficiation, excessive moisture may lead to screen and chute clogging. In transportation, moisture values above transportable moisture limit may cause cargo instabilities, especially in regard to vessels. Moisture is a non-stationary variable that depends on spatial and time distributions. Therefore, classical estimate methods such as ordinary kriging are not appropriate to calculate moisture values. Here, we present an extension of the Normative Mineralogy Calculation to indirectly estimate moisture, considering seasonal influence. This study in based on three iron-ore mines, Galinheiro, Pico and Sapecado. They are located in the Quadrilátero Ferrífero of Minas Gerais, Brazil, a world-class iron-ore district. The method proposed herein provides useful information that can be applied elsewhere. Our results indicate that compact ores show low moisture values with little seasonal influence, while soft ores and canga (iron-rich duricrust) are strongly influenced seasonally due to higher porosity and greater capacity of retaining water in the crystal structure of minerals, such as goethite. Moisture variations may exceed 2% along the year. Such variations are enough to preclude the beneficiation of certain iron ores during the rainy season. For this reason, moisture has been regarded as an essential variable in short-term mining. Article Highlights: Moisture is a critical variable in iron-ore processing, handling and transportation. Moisture depends on spatial and time distributions; hence classical methods are not appropriate to quantitatively estimate it. This study proposes an indirectly estimate of moisture considering seasonal influence. Compact iron ores are little influenced seasonally, while soft iron ores and canga (duricrust) are strongly affected by the rainy season due to their higher porosity and greater capacity of retaining water. The seasonal effect on moisture is an essential variable that must be consider to better effectiveness of iron-ore mining sequencing and beneficiation.

Um dos grandes problemas existentes na indústria mineira são as flutuações no teor do minério, que afetam, significativamente, o desempenho das plantas de beneficiamento. Assim, faz-se necessário um controle dessa variabilidade. Para tanto, pode-se usar combinações de minério provenientes de diferentes frentes de lavra e/ou pilhas de homogeneização. As pilhas de homogeneização são caracterizadas pela sua forma, tamanho, arranjo e número de camadas. O estudo em questão se propõe a minimizar a variabilidade dos depósitos minerais, utilizando pilhas longitudinais do tipo chevron, montadas com informações de teores do modelo de blocos gerados a partir de simulação geoestatística. As análises são feitas simulando diferentes modelos de pilhas, a partir de combinações de massas e diferentes números de camadas. Os resultados demonstram uma redução de até 90% da variabilidade, possibilitando um dimensionamento mais adequado do pátio e dos equipamentos de construção da pilha.

Geostatistical simulation comprises a variety of techniques which can help on the decision-making process for uncertainties. They allow the uncertainty assessment of function responses (which depend on the simulated inputs) commonly through a non-linear relationship (net present value, interest tax return, geometallurgical ore recovery...). However, one of their limitations is that running the simulation can take considerable processing time to be executed in large deposits or large grids. Herein is presented an attempt to solve this problem in short-term modeling cases, via the use of Multiple Random Walk Simulation. This algorithm combines kriging with the simulation of independent random walks in order to generate simulated scenarios much faster than via traditional simulation algorithms. A case study is presented to illustrate the application of the method in an iron mine. The Multiple Random Walk Simulation models were properly built, respecting the reproduction of both histogram and variograms. Also, the speed-up was compared with standard methods of geostatistical simulation and there was a considerable speed gain with Multiple Random Walk Simulation (3.39 to 5.65 times faster than the others).